Tim Cavanaugh has a devastating cite from just-retired Obama economic adviser Christina Romer:

The generally precise Romer spells out the difference for us: Using this approach, the estimated multiplier for monetary policy is 0.823 and the estimated multiplier for fiscal policy is -0.233.

You don't say. Gee, that would have been nice to know a few trillion dollars ago.

Democrat Party water-carriers like Paul Krugman love Keynesian economics, with its assumed large fiscal multiplier, because it meshes so perfectly with leftism's general preferences: more government, bigger government, more public-sector employees, higher pay for those employees -- and, naturally, higher taxes to go with all that. Their continued insistence we need to spend (and tax!) more, more, more even as unemployment goes higher and deficits mushroomm is growing ever less credible with each additional "unexpected" signal of economic failure.

If there's one positive to come out of the Great Recession, it should be the end of Keynesian economics as a serious policy choice. The notion you can grow the economy via North Korea-style command economics should have been long-dead even before Romer's 1992 paper, but Obama's miserable failure may finally drive a stake through this productivity-sucking, economy-killing meme.

Let me put this simply -- and contradict a too-widely-held assumption of macroeconomics:

THERE IS NO SUCH THING AS AGGREGATE DEMAND.

Government spending is not demand, it is command spending (at least for purposes of stimulating the economy). To "aggregate" it with private sector demand is like counting your dog's ringworm as a "pet" on a census form. It does not follow the same rules as private sector spending, as it is always seized and distributed according to law/fiat by bureaucrats indifferent to costs and benefits, not exchanged consensually between self-interested private parties seeking to maximize their utility. That's why Keynesianism is "unexpectedly" falling flat on its face before our eyes: it relies on a fallacious aggregation.

(Now, generally about this time someone on the left perks up and says "But... roads!" Yes, we need roads (but tend to also get Bridges To Nowhere), just as we need regulation, laws, etc. My point is not that all government spending is wasteful, just that government is 1) not very good at allocating resources to what is actually useful versus wasteful/cronyist/populist, and 2) is spending far more than it should, esp. given #1. Optimal government spending levels (i.e. those associated with highest levels of economic growth) appear to be something between 20% and 30% of GDP, whereas in the U.S. those levels are now exceeding 40%.)